Multimodal Knowledge Base
Dify now supports image vectorization and cross-modal retrieval, enabling text-to-image, image-to-text, and image-to-image search.
RAG goes beyond text-only search, so diagrams, screenshots, and photos can be understood and used by AI together with text.
Critical Security Update
Includes CVE fixes for React / Next.js and performance improvements for Excel and DOCX processing.
Strongly recommended upgrade to avoid potential security and stability risks.
For more details:
